The Gong Show

Here’s a business opportunity for anyone looking for an idea.  Many universities are making free courses available online via a movement called OpenCourseWare.  The course materials being offered for free online from many of the leading academic institutions of the world is an amazing data asset. Unfortunately, there’s not much community services (blogs, wikis, forums, etc) baked into these university sites, and the only organization that ties these offerings together today (the OpenCourseWare Consortum) is dropping the ball on creating a meta-layer that adds real value.  I think there’s an opportunity for a web service to come into this market and provide:

• An index of all courses available that is searchable/sortable by discipline, difficulty, type of media available, activity, and school.  Today, the only way to find courses is to first choose a school, and then browse by discipline… seems silly to me and needs obvious fixing.
  • This index would ideally be driven by a wiki, so people could edit the data set that drives the index collectively as more universities come online with courses (or as universities remove courses).
• User-generated reviews of course quality.
• A forums or chatrooms for people taking the same class together.  Learning can be more difficult for certain people when done alone with a book. It would be great to create a way for people to work together and help each other with the classes that interest them.
• Tie into Meetup so people can meet around their geography to discuss a course they’re all taking together.
• Accessible via Facebook App so that people can leverage the group organization resources of Facebook to consume, rate, and enhance the OpenCourseWare offerings.

The best organization positioned to do this today is the OpenCourseWare Consortium, but their website today is hardly more than brochure-ware.  They offer a list of all participating school in the OpenCourseWare project, but their members page is a static website that links only to the homepages of the universities instead of the OpenCourseWare index for each university. It makes finding actualy OpenCourseWare content a real pain, which means the only reasonable resource for finding OpenCourseWare materials is Google. This post could be considered the outline of a new company, or my dream product roadmap for the OpenCourseWare Consortium.

The internet decided that today we should all talk about our first job, so here goes. To be honest, I don’t remember which came first, but it was one of the following two jobs:

• Caddying at the local country club
• A garden weeding service I started.  I posted fliers around the neighborhood got people to hire me to do weeding at an hourly rate.

If the money were a bit better, I wouldn’t mind going back to caddying.  I loved being outdoors.  I loved watching good golfers play. And, I’d always take a monumental nap once I got home. One of my favorite books growing up was The Green Road Home, which was a book about a writer who spent some time caddying for Brad Faxon on the PGA. I could totally see doing that. Too bad pro tour bags weigh as much as I do.

Google released Q4 financials this week, and the general response in the media has been enthusiastic.  They didn’t knock the ball out of the park, but they beat street estimates.  Alley Insider has a concise summary of the key financial details:

—Net revenue: $4.2 billion vs. estimates of $4.18 billion.
—Non-GAAP EPS $5.10 vs. $4.96 per share.
—Google sites revenue up 22%
—Headcount: Company added only 99 employees in the quarter vs. estimates of 500 new employees.
—Paid Click Growth 18% vs. 17% estimates
— Total cash pile now $15.85 billion.

I want to focus specifically on paid click growth because I see paid click as an essential core metric that provides excellent insight into the health of Google’s business.  According to the call transcript, paid click growth was up 18% y/y and 10% q/q. (q/q = quarter over quarter). Paid clicks can be calculated as follows:

(ads displayed) * (click through rate) = (paid clicks)

… and, “ads displayed” is a function of pageviews times average ads per page, so lets unpack this another level…

(pageviews) * (avg ads displayed / page) * (click through rate) = (paid clicks)

So, Google has three dials they can control in order to increase paid clicks.  Hypothetically, if paid clicks grew 10% q/q because Google increased the “avg ads displayed / page” by 10% then the logical conclusion would be that pageviews were flat and the click through rate was flat.  That means to me that Google’s core traffic and ad quality did not grow at all… instead, Google simply started monetizing their pageviews more aggressively.

Remember, that last scenario was hypothetical… so what actually happened? According to AdGooRoo (via AlleyInsider), Google increased the average ads displayed / page by 50% q/q. Additionally, according to comScore, pageviews at Google increased by 5.7% q/q. Now, clearly a random company (with a name as bad as “AdGooRoo”) is not the most trusted source in the world, but even if that estimate of a 50% increase is off by a factor of 5x (or 500% overestimate on a 10% increase in ads displayed / page), then the click through rate on an ad still fell substantially in order to offset the increase in pageviews q/q.

My conclusions? The click through rate on Google ads likely fell q/q.  This could be due to a number of potential factors:

• Google’s users are becoming more savvy and are gradually becoming less likely to waste their time with and ad.
• Google is now more willing to run more marginal (less relevant) ads that result in a lower avg click through rate.
• A user interface change to the look and feel of adwords/adsense resulted in lower click through rate.

I’m sure there are other possible factors that could lead to this conclusion.  Furthermore, it’s possible that AdGooRoo’s evidence is total garbage, and in fact ads displayed / page only rose ~5%, along with pageviews, which would account for the 10% increase in paid clicks q/q. But, I have heard other anecdotal evidence in casual conversation to support the fact that Google has been juicing up ads / page, so the AdGooRoo study was not that surprising to me.

Disclosure: I’m long GOOG, and will continue to be despite this analysis.

msg:

Twitter presents their audience with a challenge - if you don’t like what we’ve done, you have every means to make it better for free, so don’t complain, do something.

Originally posted as a comment by gbattle on betaworks using Disqus.

For nearly every other company, the comsumers’ response to the “Twitter Challenge” is to give up, walk away, and never think twice. Other companies have equally open APIs (many inspired by Twitter), yet why is the consumer response in the case of Twitter’s problems, “Man, this is broken, I should spend 30 hours hacking their API to scratch my itch”?  They’re doing something right, but it’s hard to pinpoint exactly what it is.